我正在阅读 Ilya Sutskever 和 Quoc Le 的基石论文Sequence to Sequence Learning with Neural Networks。在第一页，它简要提到：

A surprising example of the power of DNNs is their ability to sort
N N-bit numbers using only 2 hidden layers of quadratic size 

谁能简要概述如何仅使用 2 个隐藏层对数字进行排序？